No abstract

The author here shares his vision of next-generation models for seismic soil-structure interaction (SSI) analysis. These models should combine reasonable considerations of wave effects in half-infinite soil with a correct representation of nonlinearity in the structure, and in both the so-called near field, i.e., in that part of soil near a base mat, and in the soil-structure contact surface. The far field, i.e., all of the soil except for the near field, is treated as a linear horizontally layered medium, as is currently done in the well-known program SASSI. The importance of considering nonlinear effects even in very stiff structures like NPPs was shown by the March 2011 Great East Japan Earthquake that hit northeastern Japan’s Pacific coast. Although the idea of calculating SSI wave effects in the time domain has been around for several decades ago, current NPP design practices are linear. Next-generation SSI models should enable practical time-domain analysis. The author suggests a road map – the sequence of problems to be solved to achieve a proposed level. Some of these problems have already been solved, at least in principle, but other solutions are yet to be found. The author describes the current status of his research and ideas about implementing modern computational techniques such as parallel computation.

No abstract